Hacking wine - is it possible?

Update, June 2016: Read our most recent post on why the future of food is synthetic:

The future of food is synthetic.

I made a post on reddit in March attempting to create wine on a compound-by-compound basis. The idea is to make wine by simply mixing raw ingredients together. Why? I’m curious, I can’t afford the best vintage wines, and this seems fun. Anyway, a number of people seemed interested, so here’s how I did it:

Ingredients

I grabbed these supplies from my local brew store and safeway:

• Tartaric acid
• Malic acid
• Tannin powder
• Vegetable glycerin
• Ethanol
• Sucrose

Tartaric acid and malic acid for sourness and acidity; Tannin for some astringency; vegetable glycerin for mouthfeel and consistency; sucrose for a little bit of sweetness; and of course, ethanol for me to get drunk.

For flavor compounds (aka volatile organic compounds, VOCs, or aroma compounds), I had these on hand from work:

• Ethyl hexanoate (smells like pineapple)
• Butanoate (strong scent of grape juice)
• Limonene (citrus/lime)
• Acetoin (rich butter smell — like popcorn at the movies)

Ideally I wanted a lot more flavors to experiment with, but figured these would do for a first pass. It sounded like I could make a white wine, probably along the lines of a riesling, a chardonnay, a grigio, or a moscato d’asti.

I had no clue what the right amount to use for each ingredient. I referenced a couple of papers and went with the median values. Interestingly, some flavor compounds were reported at 10ug/L while others at 150,000ug/L. If I was making wine the size of an olympic pool, that’s the difference between adding a drop versus adding a bucket into the water.

Results

I separated the ingredients into two groups: 1) The Base (which consisted of acids, sugars, alcohol, and glycerin) and 2) The Flavors (which consisted of VOCs). For reference I had a bottle of 2014 Kendall-Jackson Chardonnay.

Making the base was simple, I just weighed and mixed the ingredients in water. I upped the levels of tartaric and malic acid, 25% at a time, until I was satisfied with the overall tartness. I couldn’t discern a difference in taste between the two acids — they seemed equally sour. I added too much tannin the first time and it made my base unbearably bitter, so I toned it down by half. Tannins also gave my mix a nice golden color.

Glycerin had a more positive effect than I expected. It made my base “smoother” and gave it a more “rounded” taste. Since I can’t make you taste it right now, here’s a gif I made that kinda shows what I mean.

Pay attention to the right side of each glass. Notice how adding glycerin slows down the fall of the liquid.

At this point, glycerin also gave my base a tiny bit of sweetness — I thought that was enough, so I didn’t add sucrose. I poured enough ethanol to make a 12% ABV solution, then topped up the base to 750mL with water. The base by itself tasted acceptable — it was crisp, the tannin gave it dryness, and the mouthfeel was heavy and rounded.

Next, it was time to add the flavors. Getting the “right” flavor profile was a lot more challenging than expected. The biggest problem was finding the correct proportions of VOCs. I paid attention to three things: 1) the intensity of each VOC, 2) how VOCs affected each other’s flavors (ie. would pineapple negate or amplify the grape?), and 3) the interplay of flavor between VOCs and the Base. My aim was to get in the same ballpark (of intensity and complexity) as my reference wine.

Adding 5 drops of butanoate gave me a subtle grape taste. With 50 drops, I just made Welch’s grape juice. The ideal amount is likely between the two. Hexanoate was a similar story. Limonene gave me distinct citrus aroma and reminded me of some german rieslings, particularly Rudolf Muller. These first 3 VOCs worked in synergy and gave my wine a fresh somewhat-tropical taste (which was not surprising at all). My wine still lacked the richness of chardonnay, so I added acetoin. This was so pungent that I decided to add the lowest reported value I could find. It was still too much and I ended up with something like melted butter, which was just bad.

The base affects only the taste, while the VOCs affect both taste and smell. All the formulations lacked richness. Overall taste is flat. Doubling the amounts of VOCs helped improve the flavor as you first drink the wine, but the intensity tapers quickly. There is a distinct layering of flavors — as the wine enters the mouth, first, you taste the VOCs (2–3 seconds), then you get some sort of alcohol taste. There is no after taste.

Conclusion

In the end, I wasn’t 100% happy with the wines I made. I spent the whole weekend making ~15 different formulations, and I’m not much closer to the taste of chardonnay or any other wine (though I’m not discouraged). It’s not a good wine yet, but it was acceptable enough to drink. I think two fundamental questions remain: 1) what is the optimal amount for each ingredient, and 2) what ingredients are missing?

Ultimately, there isn’t a right or wrong selection of flavors here, it just depends on what kind of notes you want to taste. I actually found this quite profound — that, for the first time, the taste of the wine I’m drinking is completely up to me.

I expected that this would be hard. And it’s challenging enough that I’d like to keep going to see how far I get. I think this first experiment shows that 1) there is the possibility of making wine on a compound-by-compound basis, and 2) that the ingredients are readily available and quite easy to use.

I’ll post here once or twice a week, as I get more results. I promise to make these way shorter next time. I didn’t include a lot of numbers in this update, but if anyone’s interested in trying this out, just let me know! It’ll be great to share ideas and data.

Thanks for reading! Let me know your thoughts :)

Here’s what I plan to do next:

• On Base ingredients: I’ll be doing a matrix grid to find optimal amounts for each ingredient. I’ve barely grazed the surface! There are a lot more ingredients to add — fructose, proteins, complex carbs, etc.
• On identifying compounds/ingredients to use: My friend is running two different wines on GC for me. I should have the results by the end of the week.
• On flatness: I think increasing with the levels of acids can solve this. Theres also the case for adding lactic acid.
• On lack of longevity and after-taste: Increasing the VOCs and/or modifying/adding ingredients to the base may improve this.
• On layering of flavors: Maybe an alternate source of ethanol will have less or no alcohol taste.

Extra details:

• All ingredients used are food-grade and food-safe
• I had to adjust the pH of my Base with some potassium hydroxide to match the reference wine.

(This was originally posted on Reddit on March 2016)

4/23/16: I’ve received a couple of requests to re-create a Dom Perignon. Here’s the link if you want to get one! (use code INDIEBIO to get it half off)